This package provides a piped query generator based on Edgar F. Codd's relational algebra, and on production experience using SQL and dplyr at big data scale. The design represents an attempt to make SQL more teachable by denoting composition by a sequential pipeline notation instead of nested queries or functions. The implementation delivers reliable high performance data processing on large data systems such as Spark', databases, and data.table'. Package features include: data processing trees or pipelines as observable objects (able to report both columns produced and columns used), optimized SQL generation as an explicit user visible table modeling step, plus explicit query reasoning and checking.

This package provides a reliable and validated tool that captures detailed risk metrics such as R CMD check, test coverage, traceability matrix, documentation, dependencies, reverse dependencies, suggested dependency analysis, repository data, and enhanced reporting for R packages that are local or stored on remote repositories such as GitHub, CRAN, and Bioconductor.

Adds menu items for discrete choice experiments (DCEs) to the R Commander. DCE is a question-based survey method that designs various combinations (profiles) of attribute levels using the experimental designs, asks respondents to select the most preferred profile in each choice set, and then measures preferences for the attribute levels by analyzing the responses. For details on DCEs, refer to Louviere et al. (2000) <doi:10.1017/CBO9780511753831>.

R^2 measure of explained variation under the semiparametric additive hazards model is estimated. The measure can be used as a measure of predictive capability and therefore it can be adopted in model selection process. Rava, D. and Xu, R. (2020) <arXiv:2003.09460>.

Combined with RRphylo', this package provides a powerful tool to analyse and visualise 3d models (surfaces and meshes) in a phylogenetically explicit context (Melchionna et al., 2024 <doi:10.1038/s42003-024-06710-8>).

This package provides a thin wrapper around the tiktoken-rs crate, allowing to encode text into Byte-Pair-Encoding (BPE) tokens and decode tokens back to text. This is useful to understand how Large Language Models (LLMs) perceive text.

Allows loading and displaying an Observable notebook (online JavaScript notebooks powered by <https://observablehq.com>) as an HTML Widget in an R session, shiny application or rmarkdown document.

This package provides a common framework for calculating distance matrices.

Designed to be compatible with the R package DBI (Database Interface) when connecting to Amazon Web Service ('AWS') Athena <https://aws.amazon.com/athena/>. To do this Python Boto3 Software Development Kit ('SDK') <https://boto3.amazonaws.com/v1/documentation/api/latest/index.html> is used as a driver.

This provides a robust estimator for stochastic frontier models, employing the Minimum Density Power Divergence Estimator (MDPDE) for enhanced robustness against outliers. Additionally, it includes a function to recommend the optimal tuning parameter, alpha, which controls the robustness of the MDPDE. The methods implemented in this package are based on Song et al. (2017) <doi:10.1016/j.csda.2016.08.005>.

This package provides functions to perform robust stepwise split regularized regression. The approach first uses a robust stepwise algorithm to split the variables into the models of an ensemble. An adaptive robust regularized estimator is then applied to each subset of predictors in the models of an ensemble.

This package provides methods for ranking responses of a single response question or a multiple response question are described in the two papers: 1. Wang, H. (2008). Ranking Responses in Multiple-Choice Questions. Journal of Applied Statistics, 35, 465-474. <DOI:10.1080/02664760801924533> 2. Wang, H. and Huang, W. H. (2014). Bayesian Ranking Responses in Multiple Response Questions. Journal of the Royal Statistical Society: Series A (Statistics in Society), 177, 191-208. <DOI:10.1111/rssa.12009>.

Linear regression functions using Huber and bisquare psi functions. Optimal weights are calculated using IRLS algorithm.

Evaluates R source codes by variable and/or functions names. Similar source codes should deliver similarity coefficients near one. Since neither the frequency nor the order of the used names is considered, a manual inspection of the R source code is required to check for similarity. Possible use cases include detection of code clones for improving software quality and of plagiarism amongst students assignments.

This package provides functions for fitting a linear regression model with ARIMA errors using a filtered tau-estimate. The methodology is described in Maronna et al (2017, ISBN:9781119214687).

This package provides a programmatic interface to the Species+ <https://speciesplus.net/> database via the Species+/CITES Checklist API <https://api.speciesplus.net/>.

Wrapper for Datamuse API to find rhyming and other associated words. This includes words of similar meaning, spelling, or other related words. Learn more about the Datamuse API here <https://www.datamuse.com/api/>.

Traditional latent variable models assume that the population is homogeneous, meaning that all individuals in the population are assumed to have the same latent structure. However, this assumption is often violated in practice given that individuals may differ in their age, gender, socioeconomic status, and other factors that can affect their latent structure. The robust expectation maximization (REM) algorithm is a statistical method for estimating the parameters of a latent variable model in the presence of population heterogeneity as recommended by Nieser & Cochran (2023) <doi:10.1037/met0000413>. The REM algorithm is based on the expectation-maximization (EM) algorithm, but it allows for the case when all the data are generated by the assumed data generating model.

Peaks Over Threshold (POT) or methode du renouvellement'. The distribution for the excesses can be chosen, and heterogeneous data (including historical data or block data) can be used in a Maximum-Likelihood framework.

Allows users to easily create references to R objects then dereference when needed or modify in place without using reference classes, environments, or active bindings as workarounds. Users can also create expression references that allow subsets of any object to be referenced or expressions containing references to multiple objects.

An implementation of the QUEFTS (Quantitative Evaluation of the Native Fertility of Tropical Soils) model. The model (1) estimates native nutrient (N, P, K) supply of soils from a few soil chemical properties; and (2) computes crop yield given that supply, crop parameters, fertilizer application, and crop attainable yield. See Janssen et al. (1990) <doi:10.1016/0016-7061(90)90021-Z> for the technical details and Sattari et al. (2014) <doi:10.1016/j.fcr.2013.12.005> for a recent evaluation and improvements.

Download and open manifest files provided by the Copernicus Global Land Service data <https://land.copernicus.eu/global/>. The manifest files are available at: <https://land.copernicus.vgt.vito.be/manifest/>. Also see: <https://land.copernicus.eu/global/access/>. Before you can download the data, you will first need to register to create a username and password.

Ranking of Alternatives through Functional mapping of criterion sub-intervals into a Single Interval Method is designed to perform multi-criteria decision-making (MCDM), developed by Mališa Žižovic in 2020 (<doi:10.3390/math8061015>). It calculates the final sorted rankings based on a decision matrix where rows represent alternatives and columns represent criteria. The method uses: - A numeric vector of weights for each criterion (the sum of weights must be 1). - A numeric vector of ideal values for each criterion. - A numeric vector of anti-ideal values for each criterion. - Numeric values representing the extent to which the ideal value is preferred over the anti-ideal value, and the extent to which the anti-ideal value is considered worse. The function standardizes the decision matrix, normalizes the data, applies weights, and returns the final sorted rankings.

Randomization inference procedures for simple and complex randomized designs, including multi-armed trials, as described in Gerber and Green (2012, ISBN: 978-0393979954). Users formally describe their randomization procedure and test statistic. The randomization distribution of the test statistic under some null hypothesis is efficiently simulated.